Disable ambiguous attribution paths for cursoids with multiple PMX campaigns

When multiple PMX campaigns share the same cursoid (e.g. PMX + Reset),
paths 4 (cursoid+utm_source) and 5 (cursoid+UserAgent) are disabled for
all of them to prevent double-counting Airtable leads. Only direct paths
(GACampaignID, GoogleCampaignID, attr_referer) are used in those cases.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

airtable_client.py

@@ -336,7 +336,7 @@ class AirtableClient:
         for i in range(0, len(cat_batch), 10):
             self.campaigns.batch_update(cat_batch[i:i+10])
 
-    def get_leads_this_month_gads(self, campaign_id: str, campaign_name: str = "") -> tuple[int, list[str]]:
+    def get_leads_this_month_gads(self, campaign_id: str, campaign_name: str = "", ambiguous_cursoids: set = None) -> tuple[int, list[str]]:
         """
         Leads del mes actual para una campaña de Google Ads.
         Cubre cinco vías de atribución:
@@ -362,7 +362,9 @@ class AirtableClient:
             is_pmx_non_leadform = (
                 "pmx" in campaign_name.lower() and "_leadform" not in campaign_name.lower()
             )
-            if is_pmx_non_leadform:
+            # Paths 4 y 5 se desactivan para PMX cuando hay múltiples PMX del mismo curso
+            is_ambiguous_pmx = is_pmx_non_leadform and bool(ambiguous_cursoids and course_num in ambiguous_cursoids)
+            if is_pmx_non_leadform and not is_ambiguous_pmx:
                 utm_source = "pmx"
             elif "search" in campaign_name.lower():
                 utm_source = "google"
@@ -373,8 +375,8 @@ class AirtableClient:
                     f"AND({{attr_cursoid}}='{course_num}',"
                     f"{{attr_utm_source}}='{utm_source}')"
                 )
-            # Path 5: leadform leads en Airtable, solo para PMX (no search, no _leadform)
+            # Path 5: leadform leads en Airtable, solo para PMX único (no search, no _leadform, no ambiguo)
-            if is_pmx_non_leadform:
+            if is_pmx_non_leadform and not is_ambiguous_pmx:
                 leadform_cursoid_clause = (
                     f"AND({{attr_cursoid}}='{course_num}',"
                     f"{{UserAgent del visitante}}='Google-Ads-Notifications')"

backfill_leadform_jun8_10.py

@@ -74,12 +74,22 @@ def run():
     else:
         print("  ✓ MetricasDiarias sin cambios.")
 
+    # Cursoids con múltiples campañas PMX activas (mismo criterio que run.py)
+    pmx_count_by_cursoid: dict[str, int] = {}
+    for c in campaigns:
+        num = _course_num(c["curso"])
+        if num and "pmx" in c["curso"].lower() and "_leadform" not in c["curso"].lower():
+            pmx_count_by_cursoid[num] = pmx_count_by_cursoid.get(num, 0) + 1
+    ambiguous_pmx_cursoids = {num for num, cnt in pmx_count_by_cursoid.items() if cnt > 1}
+    if ambiguous_pmx_cursoids:
+        print(f"  ⚠️  Cursoids con múltiples PMX (paths 4/5 desactivados): {ambiguous_pmx_cursoids}")
+
     # Recalcular totales mensuales con la nueva atribución
     print("\n→ Recalculando totales mensuales (leads_lake acumulado del mes)...")
     final_leads_data = []
     for campaign in campaigns:
         cid = campaign["google_campaign_id"]
-        leads, lead_ids = at.get_leads_this_month_gads(cid, campaign["curso"])
+        leads, lead_ids = at.get_leads_this_month_gads(cid, campaign["curso"], ambiguous_pmx_cursoids)
         at.update_gacampaignmes_leads_lake(campaign["airtable_id"], lead_ids)
         final_leads_data.append({
             "airtable_id": campaign["airtable_id"],

run.py

@@ -155,6 +155,16 @@ def run():
         if "_leadform" in c["curso"].lower() and _course_num(c["curso"])
     }
 
+    # Cursoids con múltiples campañas PMX activas → paths 4/5 ambiguos, desactivar para evitar doble conteo
+    pmx_count_by_cursoid: dict[str, int] = {}
+    for c in campaigns:
+        num = _course_num(c["curso"])
+        if num and "pmx" in c["curso"].lower() and "_leadform" not in c["curso"].lower():
+            pmx_count_by_cursoid[num] = pmx_count_by_cursoid.get(num, 0) + 1
+    ambiguous_pmx_cursoids = {num for num, cnt in pmx_count_by_cursoid.items() if cnt > 1}
+    if ambiguous_pmx_cursoids:
+        print(f"  ⚠️  Cursoids con múltiples PMX (paths 4/5 desactivados): {ambiguous_pmx_cursoids}")
+
     # Precomputar conversiones Google de cada campaña leadform por número de curso
     leadform_conv_by_course: dict[str, int] = {}
     for c in campaigns:
@@ -183,7 +193,7 @@ def run():
     for campaign in campaigns:
         cid = campaign["google_campaign_id"]
 
-        leads, lead_ids = at.get_leads_this_month_gads(cid, campaign["curso"])
+        leads, lead_ids = at.get_leads_this_month_gads(cid, campaign["curso"], ambiguous_pmx_cursoids)
         at.update_gacampaignmes_leads_lake(campaign["airtable_id"], lead_ids)
 
         metrics = monthly_metrics.get(cid)